Infectivity in blood components and Cohn plasma fractions was examined in two different experimental studies: (1) high input infectivity in normal human blood that had been "spiked" with a suspension of trypsinized intact cells from a scrapie-infected hamster brain, and (2) low (indigenous) infectivity in the blood of mice inoculated with a mouse- adapted strain of human CJD. Infectivity was assayed by intracerebral inoculation of normal hamsters (scrapie or mice CJD). High levels of scrapie "spike" infectivity were recovered in all blood components, cryoprecipitate, and fraction I + II + III, with considerably less infectivity in the last two fractions. CJD-infected mice had very low levels of infectivity in buffy coat, plasma, plasma cryoprecipitate, and plasma fraction I + II + III, and no infectivity in the last two fractions. Thus, the potential risk of CJD transmission from buffy coat, plasma, factor VIII, or immune globulin (derived from the first two Cohn fractions) appears to be very low, and to be nonexistent for factor IX complex and albumin (derived from the last two fractions). A second phase of investigation of the endogenous infectivity mouse model was initiated to answer the following questions: 1) Is the presence and distribution of infectivity the same during the incubation period as during the clinical phase of disease? 2) Is the infectivity in plasma cell-associated or cell-free? 3) Is infectivity in the blood samples detectable when they are assayed by intravenous rather than intracerebral inoculation? 4) Is infectivity in Cohn fractions carried through additional processing to the final therapeutic protein concentrate? To answer these questions, we inoculated a new series of mice, from which 1) blood samples taken at various times during the incubation period were processed and are now under assay: 2) plasma samples treated with either leukodepletion filtration or high speed centrifugation are being assayed in parallel with untreated plasma; 3) buffy coat and plasma samples are being assayed using parallel intravenous and intracerebral inoculations; 4) Cohn fraction versus final product infectivity is not yet under assay, although a new group of mice has been inoculated and will be bled when sick (in about 3 months) for assays of fractions and end-products.